COGNITION TEST 2
Working Memory and Long Term Memory
- IV - Minutes in task, DV - mean cumulative words given/ animal words given high-span people do better on verbal fluency task (generate animal names) - DUAL TASK SETTING: only high-span people's performance when introduce additional task, hurt bc high span people use strategic search (categorize animals by different criterias - some complex scheme/plan used to come up with exemplars to questions) (engage with LTM and WM by use of schemas), which is disrupted by the additional task in this paradigm. Their performance is worsened more from single task to dual task switch, low span people do not suffer worsening between the two tasks (tho their performance of generated words is worse than the high span people in every condition, just most dif. when both in single task paradigm).
Overriding Phonological Similarity Effect with grapheme synaesthesia
bc letter synaesthesia will rely on different letters (rely on automatic different color of letters as the only difference between the similar sounding phonemes), they can override the phonological similarity effect - visual color info can come in, not taxed out by phono sim effect, so rely on color of letter when phono words sound similar (house, mouse, spouse). Rely on colors associated to the letters, which is what differentiates the phono similar sounding words.
Sternberg's Short Term Memory scanning task
- each stage takes time, for some stages the amount of time it takes is held constant - first: memory set already in short term memory - timer starts running when encode probe (see probe, time takes held constant), next SCAN AND COMPARISON WITH MEMORY SET ITEMS (the time it takes to do this stage of the process can tell us how we can scan our short term memory - length of this stage is affected by how large our memory set is - increase mem set, increase time of scanning phase, longer reaction time for STM scan task), binary (yes/no decision) in head (held constant), execute motor response (press yes or no, held constant, timer stops running at end of this phase).
Atkinson-Shiffrin Model
A model for describing memory in which there are three distinguishable kinds of memory (sensory, short term, long term) through which info passes in a sequential way as it is processed. - original standard theory in info-processing approach - 1st environmental input, it takes time to work with and process info - 2nd sensory memory - we select what is important, word registered brief brief moment, we manipulate into organized information, sensory registers of 6 senses are activated, but only some of a huge part of stimuli are registered - 3rd short-term memory store/temporary/working memory- we can now understand information, we use control processes (chunking) to push very important info into LTM - STM info can lead to either or both response output (required - bring info from LTM into STM (awareness)) or LTM - LTM - If memories are permanent, infinite capacity to hold, base of known/encoded information, long-term store, permanent memory store
feature comparison model
A model proposing that items are categorized by matching the item's features to category features semantics are new feature list based on concepts or cues semantic memories as feature lists semantic features - one-element properties/characteristics/attributes Types of Features: IQ test answers have things very related to a specific concept Defining Features - most defining and essential feature of specific content Characteristic Features - features that r common but not essential to the meaning of the concept
semantic network model
A model that describes units of information in long-term memory as being organized in a complex network of associations. If break semantic structure, like change month of year from progressing naturally in song to instead alphabetical order - harder process that isn't automatic anymore semantic memory involves a hierarchical structure and a process of retrieving information from that structure Characteristics: node - point in space representing a concept nodes linked by pathways, indicating associations and specific relationships Spreading activation activates related concepts P = property, isa = "is a " member of a category Denotes specific relationship between them: Red breast is property of robin P Bird concept = - Canary (ISA) bird from 1 piece, construct activated concepts, Given Robin - automatically spreads brain to other things associated with the word robin... takes you longer to think of less semantically related information, farther branched out from main concept Activation speed for a P or an ISA is directly tied to how related nodes r in a system
Brown-Peterson Task
A simple three-letter stimulus was presented to the subject (1 sec), followed by a three-digit number. Subjects were instructed firt to attend to the stimulus, then to begin counting backward by threes from the number they were shown. Ps must try to remember the trigram letters and then count backwards by 3''s from the number given (for 3-18 secs) and then recall trigram. This counting was a distracter task designed to prevent rehearsal and prove that forgetting caused by decay. IV = time spent counting backwards DV - recall of the trigram at the end of the distractor period DEPRESSINGLY HARD - shows how bad short term memory is
Classic Mnemonic Techniques
ALL METHODS REQUIRE EFFORTFUL AND ELABORATED ASSOCIATIONS OF THE ITEMS TO BE REMEMBERED Method of Loci (Memory Palace) - using a place/route you know well (like home) to provide visual/spatial structure to items you are trying to remember, provides effective cues to recall the items you have associated with that structure - place things need to remember along the way Peg word method - pre-memorized set of rhyming pairs serve as a set of mental "pegs" onto which to-be-remembered material can be "hung" (e.g, drinking game "King's Cup" - "Ace - Race, 2 - you, 3 - me (I drink) Acronyms - creating an easy to retrieve word (or sentence) in which each letter of the word (or first letter of each word in the sentence) serve as a cue for each item you need to remember - PEMDAS, ASAP
Brown Peterson Results
Accuracy of the trigram rapidly decays within 3 seconds! But this paradigm does not just have decay and retention interval effects, it also introduces interference - specifically proactive interference if the previous trials interferes with the ability to recall the present ones - recall of new info is impaired To determine trigram accuracy across intervals of distraction time (counting backwards by 3s from number presented), Ps must experience many trials (and thus many trigrams), therefore, ability to remember the trigram may also be affected by the interference from previous trials - about 50% of the time, you are unable to recall trigram after just 3 seconds of counting backwards results so bad bc trigram task is extremely hard, taxing on resources, and time counting backwards prevents ability to rehearse the trigram and acknowledge it
Poor Encoding of A-C leading to Proactive Interference
After encoding list 1 (A-B), the encoding of list 2 (A-C) may be poor due to fatigue from already doing tasks and draining some working memory resources, decreasing the ability to recall button from queen of list 2 (bc encoding was so bad)
phonological similarity effect
An effect that occurs when letters or words that sound similar are confused. For example, T and P are two similar-sounding letters that could be confused in the rehearsal in the phonological (spef. articulatory loop component), if rehearse in phonological loop, the words are still harder to differentiate bc they sound similar, they are harder to remember.
Evidence of the Difference Between short term memory and Working Memory
Baddeley and Hitch (1974) - astounding claim about STM - described memory functions in dramatic case study by Warrington and Shallice (1979) of a man with brain damage: - he had awful Short term memory - digit span only 2, mem span for words only 1, and grossly impaired performance on the Brown-Peterson Task - decay so much faster than ususal - BUT this patient and others like him had no general impairment in their ability to learn, comprehend, and remember long 6-7 word sentences - this is working memory capacity, as it is more complex things suggested that STM and WM could not be the same thign when patients with defective STM performance exhibits no memory deficiencies in other tasks that (were at the time) attributed to STM. Baddeley: suggests that STM is just one component of a larger and more elaborate working memory
Deep Levels of Processing Fergus Craik
Both elaborative encoding and processing forms Item-Specific Processing: not associations between words, focus on create individual focus on each word themselves on a scale of 1-10 how pleasant is each word? (semantic utilization, connections more deep, increase likelihood remembe, stronger memories) Create a mental image of each word Insert Deleted Letters (e.g., C _ I N) Relational Processing: Sort the words into categories based on the words meanings and associations - chunks, clusters of words together , also semantics Create a story using all the words in the list
Pair-Specific Interference Effects
Compare accuracy for RI/PI and RI/PI CONTROL - if mem for RI/PI control is better, then RI/PI pair, then PSIE occurred When recall of A-B (list 1) is worse than recall of retroactive interference control pairs, it means retroactive interference occurred When recall of (A-C) is worse than recall of proactive interference control pairs, it means proactive intereference occurred How many pairs of each do you need to demonstrate - would require same experimental methods to control?
Release from Proactive Interference Conditions
Control = recalled some type of info (each trial only words), interference built up, no switch of info type on the 4th trial - as had more trials, build up of interference, old 3 units of info made more difficult to encode other ones each trial = 3 words, count backwards, recall 3 words Experimental - only difference is switch to number fourth trial, no more (response info?), release from/reversal of proactive interference, allows memory to bounce back - highest point is at last (4th) dot and at 1st dot on graph (bc no other previous info presented yet to do proactive interference, no old info to interfere with new yet) - mem accuracy recovers
Baddeley's Model of Working Memory
Each system helps out the central executive system All systems use attention, including the central executive - WM is a general pool of resources that is limited in capacity, resources can be used up in one store, will depend on capabilities in other store to counteract usage in one Central Executive is main part, arrows going in and out of it from the Phonological Loop (phonological loop back and forth arrow to language to episodic LTM to visual semantics to visuo-spatial sketchpad) and the Visuo-Spatial Sketchpad (back and forth arrow to visual semantics to episodic LTM to language to Phonological Loop to back and forth arrow to Central Executive) We use the central executive for WM tasks mainly, but by using it we are utilizing info that we are holding onto in lower-level areas of processing (visuo-spatial sketch-pad, phonological loop, etc.) We have smaller pods of attention for lower level processing areas.
Long Term Memory - Declarative Memory (explicit) (conscious) aware of
Episodic Memory/Events - video taped eps of what we can remember, what we did Semantic Memory (Facts) - what is dinner? In LTM, questions posed to us access different stores
Retroactive/Proactive Interference Control Pairs
Ex: in list 1, beer-shell (no shared characteristics of queen-hut nor queen-button). In list 2, soul - planet (no shared characteristics of queen-button nor queen-hut). lack interference due to none of the cues or targets sharing any features with the A-B/A-C interference pairs (bc list 1 doesn't come to mind at list 2 pairing bc not using shared features), no reason for unlearning/inhibitory suppression of Retroactive Interference Pairs, and FATIGUE/POOR ENCODING EQUATED FOR A-C AND PROACTIVE CONTROL PAIRS - serve as controlling for the effect of fatigue/poor encoding by comparing scores of the two conditions. control is enacted to test for Pair-Specific Interference Effects
Other things affecting LTM
Frequency, Distinctiveness, Depth of Processing (level of processing effects), Retrieval Difficulty (Desirable difficulties and Spacing Effect) Testing - (Testing Effect and Test-Potentiated Learning Effect)
Unlearning/Inhibitory Suppression of A-B lead to Retroactive Interference
If new presence of list 2 (queen-button) brings up list 1 (queen-hut) in mind (Due to shared cue, strong inhibition of the list 1 information may be performed to focus on list 2, then you unlearn list 1 information (queen - hut) - forgotten - retroactive interference - old info recall is interfered with by the new item
Results of Dual-Task Reasoning Task
Increased reaction time in the group where must say and try to suppress random digits, the random digits override ability to read out AB and read out and decide whether statements of (A is not preceded by B) is true or not, Negative Passive condition with random digits has the highest reaction time (longest time to respond), shows taking up attentional resources, shows articulatory suppression overriding resources from the phonological loop, so it has to borrow resources from the central executive system to do the articulation, so articulatory suppresion to engage in recognition task of AB leads to increased reaction time, increased draining of resources, overrides ability to process information (AB and AB statements). All other groups mostly parallel to each other, with lowest RT at active affirmative and increasing RT to Negative Passive - other conditions: control (lowest RT), "the the the" simplest experimental 2nd lowest, "one two three" 2nd most simple, 2nd highest RT.. random digits the HARDEST by far
Desirable Difficulties in Retrieval (Bjork)
Info in retrieval must be difficult, but not impossible, for best results of learning - retrieval will modify the memory itself Technique to speed up learning: this case REFUTES the early assumption that the level of performance attained during initial (early) learning (encoding) will be maintained over the long run, because performance right after initial learning is not a good indicator of the amount of material that was actually learned and trasnferred to LTM, performance right after is being erroneously boosted by receny effect, the assumption was that the moment we obtained the skill we gained a higher level of performance, but it really does not say anything about long term performance in reality performance during initial tests reflects momentary accessibility of the knowledge and NOT HOW WELL the info has been stored in memory for the long term Bjork: in order for a person to learn and remember info for a long time, learning must be challenging to a desirably difficult degree in order for memory modificaiton (strengthened memory) to occur - must be adequately challenging for the mem to go to LTM
Cued-Recall (Easiest)
Like a fill in the blank test, easiest because the cue can help guide us to the retrieval pathway/ help the retrieval, but primacy and recency effects are still at play without controls - list-pairs utilizes a given cue, such as when given the first word for paired-word test, cue gives you help to retrieve other target word
Shadowing Task (Conway, Cowan, and Bunting (2001))
Low memory span: 65% detect name in dichotic listening task - cocktail party effect High memory span: only 20% detect High-span people more effective in selectively listening to shadowed message - more able to focus on the task given to them, less likely to be distracted
Free Recall
Only report the item, not the order of presentation/position of the items. We see both primacy and recency effects (tho can control for one of them to make the task harder). This, along with serial recall, pulls info from short term memory without a cue to guide brain to the retrieval pathway/memory. Primacy - occurs bc there is more time to rehearse those items before recall, thus more likely to be transferred to Long Term Memory. Increasing trial durations = more opportunity, time to rehearse, increase primacy effects (Words at start of list remembered better) Recency is explained simply by the task employing immediate recall (less effect of decay than the middle, easier to do), thus, if trials durations are increased, there is no effect on the recency effect, because no matter how long trials are, still going to say the last things you heard, the things you just heard at the end (there is still an end of the trial no matter how long the trials are), not enough time has gone by for words at end to decay still.
Frequency Ebbinghaus Research: Savings Score
Original Learning Score = recording of how many times (trials) you had to repeat the list to get it 100% correct (list of items learned) Re-Learning Score = some time later, re-learn the list, record how many trials it now took to get 100% correct Next time you must relearn and retrieve and create memory again, it will take less time (Original Learning Score - Relearning Score) _______________________________________________________________ = Savings Score Original Learning Score Savings Score = % reduction in the number of trials necessary for re-learning (savings... of time./trials needed.. score) Repetition increases savings score (larger reduction in # of trials required for re-learning) - implies FREQUENCY is a fundamental variable in learning and remembering - ALSO REHEARSAL is a fundamental variable Increasing time between original learning and re-learning = decreased savings score bc of forgetting over time, bc of decay over time of memory - less time in between original and re-learning, better savings score Ebbinghaus' Classic Forgetting Curve = if no time pass (RI is zero), 100% savings score, immediately info in STM and test right in front of you, immediacy (recency) effect RI 20 mins = significant drop in recall, as increase time, curve flattens out
Sternberg's Short Term Memory Scanning Task - Possible Outcomes for how we scan and compare items in STM revealed from the number of items in memory set and the time it takes to respond when probe was present in encoding or absent
Parallel Search - automatic finding of probe information, information "pops out", thus no effect on results from memory set size bc it just pops out no matter the size, and there is no difference between time it takes to respond when the probe is there (yes) or (no) - both lines are flat and parallel with each other for yes and no. - scan not implicated here, just pops out. Dif between reaction time on Yes and No responses, and memory set size increases the Reaction Time for both stimuli, but set size impacts yes response more (?), set size leads to big difference between reaction time for yes response and no response - no response will not vary with different set sizes, will always be same dependent on set size bc always at end of list will no response be reached. Serial, Self-Terminating Search - all items are scanned individually, until the probe in question is seen in the memory set size, end search with the arrival of the probe item in memory set, the "yes" response (when the probe is present) is affected by memory set size, the "no" is not because all items will be scanned and scan will only stop shorter when the probe is seen. The yes response is quicker than the no response bc you will get to the probe before you finish the whole set most of the time (odds probe is at very end every time is very unlikely). No means you scan every item to determine that the probe is not present. Serial Exhaustive Search - No matter whether the probe is present or not, there is no difference in reaction time between the yes and no response of whether probe is present or not - reaction time is only impacted by the number of items in the memory set. No matter how many items are in the set and whether or not the probe is present, still search every item exhaustively and individually, even if the probe was already seen in memory set. This graph shows reaction time increasing equally for yes and no response as memory set size increases. Slowest possible outcome here, but most cautious.
Metamemory: Remember vs. Know Paradigm - Recognition Paradigms
Ps are trained to dtrmn between a recollection of a past experience elicited by a specific test item (remembering) compared to a feeling of familiarity elicited by a specific test item (familiarity) Remember - can see specific instance, all-or-none process, automatic, fast acting process (EPISODIC RETRIEVAL OF THE EXPERIENCE) Familiarity/know- strength-based process, controlled, slower process - more controlled, time to think about it, make decisions "I know I studied that, it feels familiar, but I can't recall the instance"
Pair-Specific Response Competition Leading to Proactive Interference
Recall of Study A -C (list 2) is interfered with by study of A -B (list 1) - bc cue (A or Queen) is associated with 2 targets (Hut or Button), there is competition and interference between the two targets - when asked to recall the target associated to cue Queen is List 2, old information (list 1) can interfere with the recall of the newly presented item (list 2) - would likely be primacy effect, when old info is stronger rather than new info.
Easiest to Hardest Cognitive Processing
Recognition, Cued Recall, Free Recall, Serial Recall - more than one piece of task required in serial
Working Memory
Requires you to hold information in Short Term Memory while manipulate the info, and hold other info. - Solving Fraction Problem with multiple operations in numerator and denominator - PEMDAS rule held in central executive system
Desirable Difficulties Theory
Retrieval strength rapidly declines over time - use it to our advantage! Retrieving information increases storage strength - more we retrieve info, stronger info of memory formed If we can get the performance on a Q of the learned material closest to the original memory trace, then we can increase the strength of the memory Our memories change every time we retrieve that info/memory, it's like boundary expansion, our brain fills in the gap, we forget important information from decay or interference, soo.... attempt retrieval when retrieval strength is low, bc when retrieval strength low, retrieval is more difficult to do than when retrieval strength is high... this will lead to greatly increased storage strength If can be successful at retrieving at longer time lapse, further strengthen the correct info!! increase memory strength. when retrieval strength is high, retrieval attempts do not increase storage strength, bc retrieval is too easy
retrieval practice effect
Retrieving info from memory changes the memory and promotes long-term retention of the memory = TESTING EFFECT - given tests instead of restudy - it is DIRECT EFFECT ON PERFORMANCE, DIRECTLY INCREASES MEMORY FOR THAT TEST MATERIAL Recall is better for LTM than recognition bc RECOG is easier, Recall is like a longer retention interval for info in that it makes long-term lasting of memory better by being more difficulty, increasing worth of retrieval by being more weak memory to retrieve in first place Benefits from overt retrieval are greater than recognition (such that of desirable difficulties), short answer (free recall) is better than multiple choice (Recog)
Comparison and decision process in the Feature Comparison Model
SENTENCE VERIFICATION TASK: Simple "subject-predicate" sentences to verify - judge accuracy of sentence given a prior sentence process model - how long it takes to complete each task GLOBAL FEATURE COMPARISON: STAGE ONE: Feature overlap score: Low-Score = 1: "Fast no" - a robin is a bulldozer, fast no. Same extreme degree - high score feature overlap - high score= 9 : "fast yes", a robin is a bird Intermediate score of 5 = Comparison of Defining Features - STAGE II feature mismatch, score 5 = "slow no" to a bat is a bird feature match, score = 5, "slow yes" a chicken is a bird - subject - what is feature match to bird? Chicken not pedagogical representation of bird, but still overall overlap have to be considered, leads to consider longer, longer time react - look more closely, do features match or mismatch?
Sternberg Short Term Memory Scanning Task Results
Show Serial Exhaustive Search: we can items in short term memory one at a time, going through every item in the memory set size, no matter whether the probe was already seen (yes response) or not - as set size increases, Reaction TIme is affected, but there is no significant difference in Reaction Time for yes or no responses, both are a linear, diagonal line increasing in RT as set size increases. As 1 item added to the list in the encoding phase, it takes 38 milliseconds longer to scan memory and complete the response. Q: is there interference here?
Phonological Loop of Baddeley's Model of Working Memory
Speech and sounds, recycles verbal and auditory info for immediate recall and uses it two components: phonological store - holds verbal info Separate from the phonological store, the articulatory loop rehearses/refreshes verbal info in the phonological store, - is the articulatory loop in the phonological store? - translates visual info into verbal info and deposits in the phonological store?
Pair-Specific Response Competition - Retroactive Interference
Study A - B is interfered with by Study A -C. Cues are unrelated to one another, participants must form connection on their own. participants knew info would be tested on. cue associated with 2 targets (cue queen is associated with hut from list one and button with list 2) - when asked to recall the target associated with cue queen from list 1, the recall of list 1 (old information - hut) is interfered with by the processing of the new information (list 2 - button). Queen - Button so strong (prob. recency effect) that cannot recall hut from queen. Cue associated with 2 targets - show how 2 items compete and interfere with one another.
Why do mnemonics work?
THEY ENGAGE IN THE 3-STEP SEQUENCE OF HOW INFO IS LEARNED AND REMEMBERED - provides elaborated structure for acquiring information - ENCODING, taking in, 1st step - Storage/Retention - 2nd step - the act of creating associations (working memory) (visual images, rhymes) leads to durable and distinctive records of the material to-be-remembered Retrieval - bring it into awareness - provides effective cues for recall - 3rd step - (the cues have been formed at this point)
memory consolidation - TWO TYPES
Takes time for memories to become relatively permanent and reliable the gradual, physical process of converting new long-term memories to stable, enduring memory codes through sleeping (the neurological processes here help aid consolidation), brain flushes itself out, gets rid of bad stuff, strengthen good connections and forgets bad TWO TYPES Synaptic Consolidation: stored for up to 2 weeks (days - weeks), synaptic connections btwn neurons not strong enough to maintain info permanently Systems consolidation: stored for up to a lifetime, whole systems of processing strengthen, increase memories here for a life time Organization in Storage: Memories' better organization = stored and retrieved better Organization (Clustering) often occurs automatically : when free recall task to recall fruits, animals, first names, you will say the fruits together.... the first names together... etc..
Visuo-spatial sketchpad (VSS)
The component of the WMM that processes visual and spatial information in a mental space often called our 'inner eye' maintains/holds visual info in short duration buffer, ability to mentally rotate objects,
Distinctiveness
The phenomenon in memory that states we are better able to remember information if it is distinctive or different from other information. - if this distinctiveness develops during encoding and lead to more effective retrieval cues - then it is NOT INHERENTLY WEIRD OUT OF CONTEXT, only weird in context (foot, given other items are arm, leg, and finger) Inherent Distinctiveness: arm, leg, uvula, foot - unusual item regardless of surrounding items, inherent without the need for a surrounding context and automatically captures attention, low frequency word, idiosyncratic word - like ABE control variables - like word Kumquat BOTH HELP MEMORY BY CAPTURING ATTENTION AND PROVIDING RETRIEVAL CUES
Visual Memory and Deep Processing
Visual memory is super strong, pics remembered much more than words, it supports dual-coding hypotheses (this is implicated in deep processing) Dual Coding hypothesis: words denoting concrete objects (encourages pictorial representation) can be encoded in memory twice - less than word "freedom" could be, as it is more abstract Gives two ways to retrieve the memory (word cup, image of cup) - and it is easier when word has obvious image
Release from Proactive Interference
Wickens 1972 invovled Ps in a Brown-Peterson Type Task (Short term memory), but without varying the distraction interval (count backwards by 3) time. Ps given 3 pieces of info (words or numbers) on each trial (not trigrams), after each trial they countedbackward from a number by 3s for brief period, then asked to recall the 3 pieces of info (words or numbers) When the 4th trial's three pieces of info WERE DIFFERENT than the first 3 trials (different modality, three words, then number), = RELEASE FROM PROACTIVE INTERFERENCE - recall of NEW INFO is no longer impaired by previous old information
levels of processing
a continuum of memory processing from shallow to intermediate to deep, with deeper processing producing better memory most basic level = nervous system processes incoming sensory info to organize and understand it - physical incoming sensory info higher levels = that info is processed with previous existing knowledge (LTM) to extract meaning - connect with other, familiar ideas attention can be directed toward processing basic perceptual info (shallow) or higher level meaning extraction (deep) directing attention toward a shallow level of processing disrupts deeper levels of processing, bc it still take sup attention, so can't do task that's even more draining of attention (deeper processing) while doing the shallow one - do one or the other! . if control attention away from shallow processing, helps memory bc it encourages higher level of processing
semantic memory - part of declarative memory subtype - U R CONSCIOUSLY AWARE OF IT
a network of associated facts and concepts that make up our general knowledge of the world Memory for meaning, store of conceptual knowledge created from our experience with the world, shared with community, stable across individual events = no matter who u r, 2 + 2= 4, LTM of the world not as much recollection, not as episodic Related but more implicit/nondeclarative memory: Semantic priming - ur knowledge of meaning activates in u as soon as u see something, see it automatically, auto think of cats if you see image of cats, can influence behavior on tasks... more automatic, more implicit, this specific thing less conscious of
dual task method, dual task reasoning task
a primary and secondary task are performed simultaneously, possible results: complete dependence (primary and secondary task use same mental resources, change in 1 dictates change in the other) complete independence (primary and secondary use separate mechanisms or resources - change in 1 doesn't change the other some level of dependency - primary and secondary partially share resources Dual-Task Reasoning Task - Ps shown a stimulus like "AB" then must respond "yes" or "no" to sentences about the stimulus - either affirmative (easier), active or passive (uses word is), or negative (harder), active or passive (A IS NOT preceded by B), this is the primary task, while they do this they must also enagge in either articulatory suppression (repeating #s 1-6) or repeating a random sequence of digits (hardest type of articulatory suppressison) DV - Reaction Time to saying yes or no, the statement is true or false of the image AB seen. As go thru levels, harder and harder to process
autobiographical memory - Declarative Memory
a special form of episodic memory, consisting of a person's recollections of his or her life experiences infantile amnesia - first memory usually from age 2 - 4 years old, no memory of infanthood - why? 1 . product of development of memory (from images to language based, lang gives second storage of info to increase memory) 2. begin remembering as develop self, own life evetns are consolidated bc u have life, it means something, adds context to greater life, IT IS NOT due to a lack of memory in infanthood - some memory present even before birth, we learn what foods to like in womb, know mother's voice, change in light Reminiscence Bump - superior memory for life events around age 20, things ull remember and hold onto extremely long time, maybe bc: first time events usually, and cultural expectations to remember Period of Recency - increased memory for events in life that have just happened/passed - highest point in memory, second highest at reminiscence bump, but as recency goes away for those events in period of recent, will decline and not be as strong than as reminiscence bump ones
Recall - one way to analyze the contents of people's short term memory
actively retrieving a memory, info from short term memory - 3 types: cued, free (bc without proper controls, still get primacy and recency effects when only need to recall item), serial (recall item plus position/order presented is harder than item only, item only is free recall) (increasing level of difficulty as go up list)
bottleneck metaphor
all the information you experience goes into a bottleneck and only the little information that comes out is what you concentrate on whatever gets in first, gets in limited capacity - magical number 7 +- 2, (Miller, 1956), argues you can only recall 5-9 individual units in order - but this is moderated by the speed of your mental verbalization - the faster you can spit out info from STM, the larger ur STM capacity can be - so words less memory span than for letters/digits bc takes longer to articulate in head 38ms to rehearse/scan 1 item very limited short term memory capacity of what you can hold
Metamemory
an understanding about the processes that underlie memory, which emerges and improves during middle childhood - awarness of own cognition, knowledge about own memory, can be misleading (under or over confident about having learned something) our ability to judge how good our memory is, is often very bad over conf often means worse memory than under confident
Interference
another way we forget in short term memory
Conclusions of Levels of Processing
attention can be allocated to dif levels of processing, THE TYPE OF PROCESSING (DEEP VS. SHALLOW), rather than the INTENT TO LEARN, influences what information is encoded and retained in memory so in educational settings, extraction of meaning is ultimate goal deep processing of info - in way that optimally extracts meaning - most beneficial to students
Recognition - another way to analyze the contents' of people's short term memory
binary-choice (old/new, present/absent, yes/no) and multiple-choice (3 wrong, 1 correct, recognize correct one) results assessed with Signal Detection theory - Hits, Misses, Correct Rejections, and False Alarms - old item presented, say yes it was there, this is a hit d' = sensitivity to stimulus, C = response bias (neg = liberal, pos = conservative) Primacy effect is more easy to achieve in Recognition - when primacy is strong and presumably the beginning of the list items were transferred to long term memory, then primacy effects in recognition seen. Recency effects in recognition tasks occur when the questions asking about the end of the list items in the study phase (encoding) are presented first on the test (test) - if end items occur as the first questions on test, then recency would be present, but if they occur later in the test then the first items presented on the test and recalled from memory (the act of engaging in recognition of other items first) will serve as sufficient enough distraction to wipe out the increased memory of the last presented items in encoding, wiping out the recency effect (and also no controls must be in place to wipe out recency effect, such as increased Retention Interval, distraction/filler task at end of the list presentation that won't be tested on)
Episodic Buffer - Baddeley's Model of Working Memory
binds together information from different modalities and sources chunking perceptual processes pulling episodic memory in to help remember things, make more viable chunk of info due to shared contexts, takes working memory info to long term memory Key Studies: Jefferies, Lambdon Ralph, Baddeley (2004) - engage in 2 tasks at once, decreased ability to remember, were sim words or dif words remembered more? Copeland and Radvansky (2001) - remember last word each sentence, half sentences words were phonologically similar to each other, half sentences words were phonologically dissimilar - WHOLE SENTENCE FROM LONG TERM MEMORY HELPED REMEMBER PHONOLOGICALLY SIMILAR WORDS COMPARED TO DISSIMILAR PHONO WORDS, PHONOLOG SIMILARITY INCREASES RECALL, NOT THE SAME IN PHONOLOGICAL LOOP AUDITORY STORE (which posits that decreased mem for phono similarity in auditory system)
central executive system
controls the deployment of attention, switching the focus of attention and dividing attention as needed -mainly, controls planning, initiates retrieval and decision, integrates information cent. exec applies rules (PEMDAS), retrieves rules, retrieves value and engages rules, arithmetic central executive has own limited pool of resources, resources are somewhat shared across processing domains
Optimal Spacing Factors
difficulty of the concept/skill being learned (if more difficult, use shorter spaces than average to buffer the just difficulty), how long you want to retain the knowledge (longer spacing, more LTM), spacing over greater intervals leads to memory that lasts longer
shallow level of processing - fergus Craik (1972, 1975)
encoding on a basic level based on the structure or appearance of words (how many words are in bold font? in lower-case? 5 letters? simply read the list to yourself) less connections formed, less memory
Storage and Retrieval of Episodic Memories Dependent on Context
everything happens in a context, episodic experiences occur in time and space - Godden and Baddeley : STATE DEPENDENT LEARNING - Words underwater encoding, better recall underwater, words on land encoding, better recall on land - switch contexts = worse performance, get cross-over interaction - environmental cues are encoded with the information learning in the context Encoding specificity: info encoded into memory including context ex: store word from list along with context where you read the list PHYSICAL AND INTERNAL CONTEXTS HERE
Retroactive Interference
recall of OLD info is impaired - new info is so strong that it retroactively interferes with ability to remember old info - u get new phone number and it is so important, so strong, that it causes you to forget old and own phone number
Processing Feedback - Making Correct Responses
highly important for correcting errors and maintaining correct responses = gives confidence in correct answers did get, and alerts you "cannot make that mistake again!" (Hypercorrection effect for high confidence errors, more pop out efefct than answers you gave that you weren't so sure of) Feedback influences memory Dependent on: Type of feedback: delayed feedback is better than immediate for retention, bc you hold the high confidence longer for it to have an effect but: way students process feedbakc when it is immediate vs. delayed can differ and influence the effectiveness of the feedback - maybe don't care as much about feedback, about being right or wrong, at longer delays to feeeback .. may pay more attention when it is immediate, be influenced by immediate reward of being right/or punishment.
Overcoming the Bottleneck
if you can recode (process all information together) the individual units of information into larger and more cohesive units, called chunks, you can recall more information from Short term memory STM ... but recoding info into chunks requires sufficient time and WM resources, as well as a scheme (like a template or pattern) that is well learned to do it Ex: phone numbers can be recoded into chunks that we are all familiar with (fam. with this scheme), which is a 3-4 pattern (or SSN is a 3-2-4 pattern)
Long Term Memory - Nondeclarative memory (implicit, unconscious)
includes: procedural memory: skills and habits, cooking, baking, over developed skills and habits with automaticity, spatial memory and memory navigation Implicit Memory: priming, past experiences act on our every day processing of information, it is present in the info we process in the present
short-term memory (STM)
input, rehearsal, and storage of new information observed short retentions (15-20s after encoding) when little or no info is transferred to LTM HOW WE REHEARSE INFORMATION Ex: immediate recall in digit span task STM is a SUBSET of the more elaborate WM system
The Sternberg Task
investigating how one scans the contents of their short term memory to decide if something is present or absent by implementing a paradigm where Ps are presented with list of items (#s or letters) in which THE SET SIZE VARIES ON EACH TRIAL (IV) and then, immediately after the set, Ps are presented with a probe item - Ps task is to determine if the Probe item was an item present (or absent) in the set presented (encoding phase). IV = number of the set size for each trial DV = reaction time to respond "yes" and "no" - Sternberg interested in how quick you can make decision, there is no rehearsal involved How quickly we can scan information in short term memory to come to particular response Is decay/interference at play? Not sure.
Desirable Difficulties Theory Conclusions
learning that is faster and easier is not better retrieval that is difficult promotes long term retention educators - should introduce dififculties into the learning process to promote long-term retention of the info, BUT not make learning impossible a learned must be able to OVERCOME the retrieval difficulty for storage strength to increase - difficulty must be desirable
Sternberg short term memory task limitations
limitations to conclusions: the increased reaction times across set size could still be due to parallel search in which each additional item scanned slows down the rate of scanning for all items (all done at once), parallel search possibility in which the search is just influenced by the number of items we have to search in parallel, all at once. - argue against the idea that only serial exhaustive search used Limitations to Sternberg's Process Model Assumption - Sternberg proposes that each stage is sequential and one stage must end before the next begins (assumes stages completely independent), but others (McClelland, 1979) proposed mental stages may partially overlap
Word Length Effect
longer words are harder to remember than shorter words, because increased decay and interference, it is harder to keep longer words in phonological loop and articulatory loop - also maybe take longer to verbally say in mind, so reduced recall of it???
Stroop Task: Kane and Engle (2003)
manipulate how often mismatch word w/ color presented and color and word match presented. Inhibit word, respond color: higher working memory span, more able to selectively attend to target, critical info, stroop effect: 100-ms slowdown mismatch of 50%: no difference btwn high and low span poeple mismatch of 20%: low-span people made more errors - when 80% of presented is word and color match, lower Working memory span people tend to fixate on fact that most of the time word and color match, don't care as much in memory to be cautious, pattern makes them lazy
Proactive Interference
recall of NEW info impaired - old info is interfering in forward direction with new info - old phone number of friend interferes with ability to remember new phone number of friend, so used to it Shown in Brown Peterson Task - new trigrams may be badly recalled due to the previous trigrams seen interfering with the present trigram, interference in the forward direction
working memory (WM)
more advanced, elaborate system than short-term memory the "mental workbench", place where mental effort is applied (Baddeley, 1992) like storage and manipulation of info to allow for comprehension and reasoning - all things going thru head when we remember where we have to go, what we have to do, EX: when semantics are retrieved from LTM to understand the meaning of a sentence complex operation span affects other processes measurement implicated in all IQ tests Solve problems, hold some info while we manipulate other info, daily life activities STM is a SUBSET of the more elaborate WM system
Decay
one source of forgetting, caused simply by the passage of time before testing (implication: without rehearsal, information in STM decays extremely quickly, rehearsal/processing needed to not forget and to push info into LTM long-term memory Shown in Brown-Peterson Task - about 50% of the time, unable to recall trigram after just 3 seconds of backwards counting Once decay, not able to recode or manipulate anymore bc don't know it anymore!
boundary extension
people tend to misremember more of a scene than was actually viewed, as if the boundaries of an image were extended further out. Brain fills in outter gap of photo, tend to zoom out our lense
Articulatory Suppression Effect
poor memory for verbal material if must say random words or other things while trying to rehearse verbal information in memory, it hurts memory, it drains attention and worsens performance - the words you say override the rehearsal of target words you want to remember
mental rotation - visuo-spatial sketchpad
process of imagining an object turning in three-dimensional space - ask self - does this rotated object (Second show) match the object I encoded? Same angles? Testing of visual memory If increased rotational difference between 1st object and second object (degree of angle difference - up to 120), longer Reaction Time to decide if the object is the same or not.
Serial Recall
recall items in the exact order they were presented, recency effect is typically wiped out (not there to help memory) because the recalling of the first few items is sufficient enough distraction to forget the more recent (last presented end of list) items. The primacy effect is still at play though, still can create confounds for study, but it makes task slightly easier (tho it is the most difficult recall task). THERE IS NO CUE INFO PRESENTED OT GUIDE BRAIN TO THE RIGHT MEMORY/RETRIEVAL PATHWAY - actively pulls info from memory without a cue present
Recoding/Chunking/Scheme
recoding = processing of grouping items together and then remembering these newly formed groups chunks = the newly formed groups you are remembering Scheme = grouping pattern or template
retrieval practice effect: Testing Effect (GPE)
takes advantage of generation, production, and enactment effects Generation effect: remember info better if you generate it yourself - peterson study - make own Qs of reading on quizlet than experts guided Qs Ex: reading a word v. generating a word stronger with free recall Effect larger over larger delays Production Effect: Produce information yourself - better recall Ex: say words aloud to yourself rather than reading or hearing them Work if you write or mouth it also Enactment Effect: Remember better if you perform a task - why sheltered parenting = less able kids
rehearsal
the conscious repetition of information, deliberate practicing in Short Term Memory Store, either to maintain it in consciousness (short term memory) or to encode it for storage/transfer into long term memory bring into STM, use phonological loop to rehearse, and then episodic buffer brings to LTM
Amount of Release from Proactive Interference Impacted by Degree of Similarity (Meaning, Semantics) between first 3 items and last 4th item presented
the more related the items on the 4th trial were to the first 3 trials, the smaller the release from Proactive Interference. When responses you have to make compete for each other due to similar meanings/semantics, increases interference. The release bounce back (recovery of memory) is related to how similar the changing stimuli are. Bc fruits and vegetables are closely related, there is still a good amount of interference going on, there is some release from PI, but as you go from less related items on the 4th trial compared to the first 1-3 items meaning, you get larger release from proactive interference.
long-term memory
the relatively permanent and limitless storehouse of the memory system. Includes knowledge, skills, and experiences.
spacing effect
the tendency for distributed study or practice to yield better long-term retention than is achieved through massed study or practice (cramming, often done by procrastination) makes sense with desirable difficulties / increased time lapse for retrieval strength to be lower, so retrieval is more valuable for long-term memory consolidation SPACING DURING ENCODING = BETTER MEMORY PERFORMANCE = increase time between each successive recall, increase difficulty to recall the info, so bigger boost bc retrieval is less strong at longer time intervals Kornell & Bjork (2008) - Ps studied 6 paintings from 10 artists each, in either spaced (dif artists one at a time) vs. massed (Each artist at once), then took a test on new painting by the same artists - must generalize knowledge of each artist distinct style - mass grouped - less differentaiation of styles of artists, less attention paid to it when all at once, less percent correct of new artists paintings
Representational Momentum Visuo-Spatial sketchpad
uses predictive processes to determine where objects are/will be in space - see ball rolling, test people on how far the ball went down, we over-report the distance the ball went - our visual memory goes past, overextends visual information.
Retrieval Practice: test-potentiated learning
with testing, this is present, it is ability to encode remember new (nontested) info is also better!!! INDIRECT EFFECT - testing on other info (retrieval) increases ability to learn other info!!! Successful retrieval is key - feedback for incorrect responses is IMPORTANT Makes aware how well you encoded the present info, so when same type of info is presented (new info tho), metacognitive process helps guide u to what need to do to do better to encode this test, learn this tested info TPL can be more robust than testing effect (Specific to items)
